Commit | Line | Data |
---|---|---|
1da177e4 LT |
1 | /* Generate assembler source containing symbol information |
2 | * | |
3 | * Copyright 2002 by Kai Germaschewski | |
4 | * | |
5 | * This software may be used and distributed according to the terms | |
6 | * of the GNU General Public License, incorporated herein by reference. | |
7 | * | |
8 | * Usage: nm -n vmlinux | scripts/kallsyms [--all-symbols] > symbols.S | |
9 | * | |
10 | * ChangeLog: | |
11 | * | |
12 | * (25/Aug/2004) Paulo Marques <pmarques@grupopie.com> | |
13 | * Changed the compression method from stem compression to "table lookup" | |
14 | * compression | |
15 | * | |
16 | * Table compression uses all the unused char codes on the symbols and | |
17 | * maps these to the most used substrings (tokens). For instance, it might | |
18 | * map char code 0xF7 to represent "write_" and then in every symbol where | |
19 | * "write_" appears it can be replaced by 0xF7, saving 5 bytes. | |
20 | * The used codes themselves are also placed in the table so that the | |
21 | * decompresion can work without "special cases". | |
22 | * Applied to kernel symbols, this usually produces a compression ratio | |
23 | * of about 50%. | |
24 | * | |
25 | */ | |
26 | ||
b3dbb4ec PM |
27 | #define _GNU_SOURCE |
28 | ||
1da177e4 LT |
29 | #include <stdio.h> |
30 | #include <stdlib.h> | |
31 | #include <string.h> | |
32 | #include <ctype.h> | |
33 | ||
1da177e4 LT |
34 | #define KSYM_NAME_LEN 127 |
35 | ||
1da177e4 LT |
36 | |
37 | struct sym_entry { | |
38 | unsigned long long addr; | |
b3dbb4ec | 39 | unsigned int len; |
1da177e4 LT |
40 | unsigned char *sym; |
41 | }; | |
42 | ||
43 | ||
44 | static struct sym_entry *table; | |
b3dbb4ec | 45 | static unsigned int table_size, table_cnt; |
075d6eb1 | 46 | static unsigned long long _stext, _etext, _sinittext, _einittext, _sextratext, _eextratext; |
1da177e4 | 47 | static int all_symbols = 0; |
41f11a4f | 48 | static char symbol_prefix_char = '\0'; |
1da177e4 | 49 | |
b3dbb4ec | 50 | int token_profit[0x10000]; |
1da177e4 LT |
51 | |
52 | /* the table that holds the result of the compression */ | |
b3dbb4ec | 53 | unsigned char best_table[256][2]; |
1da177e4 LT |
54 | unsigned char best_table_len[256]; |
55 | ||
56 | ||
b3dbb4ec | 57 | static void usage(void) |
1da177e4 | 58 | { |
41f11a4f | 59 | fprintf(stderr, "Usage: kallsyms [--all-symbols] [--symbol-prefix=<prefix char>] < in.map > out.S\n"); |
1da177e4 LT |
60 | exit(1); |
61 | } | |
62 | ||
63 | /* | |
64 | * This ignores the intensely annoying "mapping symbols" found | |
65 | * in ARM ELF files: $a, $t and $d. | |
66 | */ | |
b3dbb4ec | 67 | static inline int is_arm_mapping_symbol(const char *str) |
1da177e4 LT |
68 | { |
69 | return str[0] == '$' && strchr("atd", str[1]) | |
70 | && (str[2] == '\0' || str[2] == '.'); | |
71 | } | |
72 | ||
b3dbb4ec | 73 | static int read_symbol(FILE *in, struct sym_entry *s) |
1da177e4 LT |
74 | { |
75 | char str[500]; | |
b3dbb4ec | 76 | char *sym, stype; |
1da177e4 LT |
77 | int rc; |
78 | ||
b3dbb4ec | 79 | rc = fscanf(in, "%llx %c %499s\n", &s->addr, &stype, str); |
1da177e4 LT |
80 | if (rc != 3) { |
81 | if (rc != EOF) { | |
82 | /* skip line */ | |
83 | fgets(str, 500, in); | |
84 | } | |
85 | return -1; | |
86 | } | |
87 | ||
41f11a4f YS |
88 | sym = str; |
89 | /* skip prefix char */ | |
90 | if (symbol_prefix_char && str[0] == symbol_prefix_char) | |
91 | sym++; | |
92 | ||
1da177e4 | 93 | /* Ignore most absolute/undefined (?) symbols. */ |
41f11a4f | 94 | if (strcmp(sym, "_stext") == 0) |
1da177e4 | 95 | _stext = s->addr; |
41f11a4f | 96 | else if (strcmp(sym, "_etext") == 0) |
1da177e4 | 97 | _etext = s->addr; |
41f11a4f | 98 | else if (strcmp(sym, "_sinittext") == 0) |
1da177e4 | 99 | _sinittext = s->addr; |
41f11a4f | 100 | else if (strcmp(sym, "_einittext") == 0) |
1da177e4 | 101 | _einittext = s->addr; |
075d6eb1 DW |
102 | else if (strcmp(sym, "_sextratext") == 0) |
103 | _sextratext = s->addr; | |
104 | else if (strcmp(sym, "_eextratext") == 0) | |
105 | _eextratext = s->addr; | |
b3dbb4ec | 106 | else if (toupper(stype) == 'A') |
1da177e4 LT |
107 | { |
108 | /* Keep these useful absolute symbols */ | |
41f11a4f YS |
109 | if (strcmp(sym, "__kernel_syscall_via_break") && |
110 | strcmp(sym, "__kernel_syscall_via_epc") && | |
111 | strcmp(sym, "__kernel_sigtramp") && | |
112 | strcmp(sym, "__gp")) | |
1da177e4 LT |
113 | return -1; |
114 | ||
115 | } | |
b3dbb4ec | 116 | else if (toupper(stype) == 'U' || |
41f11a4f | 117 | is_arm_mapping_symbol(sym)) |
1da177e4 LT |
118 | return -1; |
119 | ||
120 | /* include the type field in the symbol name, so that it gets | |
121 | * compressed together */ | |
122 | s->len = strlen(str) + 1; | |
b3dbb4ec PM |
123 | s->sym = malloc(s->len + 1); |
124 | strcpy((char *)s->sym + 1, str); | |
125 | s->sym[0] = stype; | |
1da177e4 LT |
126 | |
127 | return 0; | |
128 | } | |
129 | ||
b3dbb4ec | 130 | static int symbol_valid(struct sym_entry *s) |
1da177e4 LT |
131 | { |
132 | /* Symbols which vary between passes. Passes 1 and 2 must have | |
133 | * identical symbol lists. The kallsyms_* symbols below are only added | |
134 | * after pass 1, they would be included in pass 2 when --all-symbols is | |
135 | * specified so exclude them to get a stable symbol list. | |
136 | */ | |
137 | static char *special_symbols[] = { | |
138 | "kallsyms_addresses", | |
139 | "kallsyms_num_syms", | |
140 | "kallsyms_names", | |
141 | "kallsyms_markers", | |
142 | "kallsyms_token_table", | |
143 | "kallsyms_token_index", | |
144 | ||
145 | /* Exclude linker generated symbols which vary between passes */ | |
146 | "_SDA_BASE_", /* ppc */ | |
147 | "_SDA2_BASE_", /* ppc */ | |
148 | NULL }; | |
149 | int i; | |
41f11a4f YS |
150 | int offset = 1; |
151 | ||
152 | /* skip prefix char */ | |
153 | if (symbol_prefix_char && *(s->sym + 1) == symbol_prefix_char) | |
154 | offset++; | |
1da177e4 LT |
155 | |
156 | /* if --all-symbols is not specified, then symbols outside the text | |
157 | * and inittext sections are discarded */ | |
158 | if (!all_symbols) { | |
159 | if ((s->addr < _stext || s->addr > _etext) | |
075d6eb1 DW |
160 | && (s->addr < _sinittext || s->addr > _einittext) |
161 | && (s->addr < _sextratext || s->addr > _eextratext)) | |
1da177e4 LT |
162 | return 0; |
163 | /* Corner case. Discard any symbols with the same value as | |
075d6eb1 DW |
164 | * _etext _einittext or _eextratext; they can move between pass |
165 | * 1 and 2 when the kallsyms data are added. If these symbols | |
166 | * move then they may get dropped in pass 2, which breaks the | |
167 | * kallsyms rules. | |
1da177e4 | 168 | */ |
61d9cdf2 M |
169 | if ((s->addr == _etext && strcmp((char*)s->sym + offset, "_etext")) || |
170 | (s->addr == _einittext && strcmp((char*)s->sym + offset, "_einittext")) || | |
171 | (s->addr == _eextratext && strcmp((char*)s->sym + offset, "_eextratext"))) | |
1da177e4 LT |
172 | return 0; |
173 | } | |
174 | ||
175 | /* Exclude symbols which vary between passes. */ | |
b3dbb4ec | 176 | if (strstr((char *)s->sym + offset, "_compiled.")) |
1da177e4 LT |
177 | return 0; |
178 | ||
179 | for (i = 0; special_symbols[i]; i++) | |
b3dbb4ec | 180 | if( strcmp((char *)s->sym + offset, special_symbols[i]) == 0 ) |
1da177e4 LT |
181 | return 0; |
182 | ||
183 | return 1; | |
184 | } | |
185 | ||
b3dbb4ec | 186 | static void read_map(FILE *in) |
1da177e4 LT |
187 | { |
188 | while (!feof(in)) { | |
b3dbb4ec PM |
189 | if (table_cnt >= table_size) { |
190 | table_size += 10000; | |
191 | table = realloc(table, sizeof(*table) * table_size); | |
1da177e4 LT |
192 | if (!table) { |
193 | fprintf(stderr, "out of memory\n"); | |
194 | exit (1); | |
195 | } | |
196 | } | |
b3dbb4ec PM |
197 | if (read_symbol(in, &table[table_cnt]) == 0) |
198 | table_cnt++; | |
1da177e4 LT |
199 | } |
200 | } | |
201 | ||
202 | static void output_label(char *label) | |
203 | { | |
41f11a4f YS |
204 | if (symbol_prefix_char) |
205 | printf(".globl %c%s\n", symbol_prefix_char, label); | |
206 | else | |
207 | printf(".globl %s\n", label); | |
1da177e4 | 208 | printf("\tALGN\n"); |
41f11a4f YS |
209 | if (symbol_prefix_char) |
210 | printf("%c%s:\n", symbol_prefix_char, label); | |
211 | else | |
212 | printf("%s:\n", label); | |
1da177e4 LT |
213 | } |
214 | ||
215 | /* uncompress a compressed symbol. When this function is called, the best table | |
216 | * might still be compressed itself, so the function needs to be recursive */ | |
217 | static int expand_symbol(unsigned char *data, int len, char *result) | |
218 | { | |
219 | int c, rlen, total=0; | |
220 | ||
221 | while (len) { | |
222 | c = *data; | |
223 | /* if the table holds a single char that is the same as the one | |
224 | * we are looking for, then end the search */ | |
225 | if (best_table[c][0]==c && best_table_len[c]==1) { | |
226 | *result++ = c; | |
227 | total++; | |
228 | } else { | |
229 | /* if not, recurse and expand */ | |
230 | rlen = expand_symbol(best_table[c], best_table_len[c], result); | |
231 | total += rlen; | |
232 | result += rlen; | |
233 | } | |
234 | data++; | |
235 | len--; | |
236 | } | |
237 | *result=0; | |
238 | ||
239 | return total; | |
240 | } | |
241 | ||
b3dbb4ec | 242 | static void write_src(void) |
1da177e4 | 243 | { |
b3dbb4ec | 244 | unsigned int i, k, off; |
1da177e4 LT |
245 | unsigned int best_idx[256]; |
246 | unsigned int *markers; | |
247 | char buf[KSYM_NAME_LEN+1]; | |
248 | ||
249 | printf("#include <asm/types.h>\n"); | |
250 | printf("#if BITS_PER_LONG == 64\n"); | |
251 | printf("#define PTR .quad\n"); | |
252 | printf("#define ALGN .align 8\n"); | |
253 | printf("#else\n"); | |
254 | printf("#define PTR .long\n"); | |
255 | printf("#define ALGN .align 4\n"); | |
256 | printf("#endif\n"); | |
257 | ||
258 | printf(".data\n"); | |
259 | ||
260 | output_label("kallsyms_addresses"); | |
b3dbb4ec PM |
261 | for (i = 0; i < table_cnt; i++) { |
262 | printf("\tPTR\t%#llx\n", table[i].addr); | |
1da177e4 LT |
263 | } |
264 | printf("\n"); | |
265 | ||
266 | output_label("kallsyms_num_syms"); | |
b3dbb4ec | 267 | printf("\tPTR\t%d\n", table_cnt); |
1da177e4 LT |
268 | printf("\n"); |
269 | ||
270 | /* table of offset markers, that give the offset in the compressed stream | |
271 | * every 256 symbols */ | |
b3dbb4ec | 272 | markers = (unsigned int *) malloc(sizeof(unsigned int) * ((table_cnt + 255) / 256)); |
1da177e4 LT |
273 | |
274 | output_label("kallsyms_names"); | |
1da177e4 | 275 | off = 0; |
b3dbb4ec PM |
276 | for (i = 0; i < table_cnt; i++) { |
277 | if ((i & 0xFF) == 0) | |
278 | markers[i >> 8] = off; | |
1da177e4 LT |
279 | |
280 | printf("\t.byte 0x%02x", table[i].len); | |
281 | for (k = 0; k < table[i].len; k++) | |
282 | printf(", 0x%02x", table[i].sym[k]); | |
283 | printf("\n"); | |
284 | ||
285 | off += table[i].len + 1; | |
1da177e4 LT |
286 | } |
287 | printf("\n"); | |
288 | ||
289 | output_label("kallsyms_markers"); | |
b3dbb4ec | 290 | for (i = 0; i < ((table_cnt + 255) >> 8); i++) |
1da177e4 LT |
291 | printf("\tPTR\t%d\n", markers[i]); |
292 | printf("\n"); | |
293 | ||
294 | free(markers); | |
295 | ||
296 | output_label("kallsyms_token_table"); | |
297 | off = 0; | |
298 | for (i = 0; i < 256; i++) { | |
299 | best_idx[i] = off; | |
b3dbb4ec | 300 | expand_symbol(best_table[i], best_table_len[i], buf); |
1da177e4 LT |
301 | printf("\t.asciz\t\"%s\"\n", buf); |
302 | off += strlen(buf) + 1; | |
303 | } | |
304 | printf("\n"); | |
305 | ||
306 | output_label("kallsyms_token_index"); | |
307 | for (i = 0; i < 256; i++) | |
308 | printf("\t.short\t%d\n", best_idx[i]); | |
309 | printf("\n"); | |
310 | } | |
311 | ||
312 | ||
313 | /* table lookup compression functions */ | |
314 | ||
1da177e4 LT |
315 | /* count all the possible tokens in a symbol */ |
316 | static void learn_symbol(unsigned char *symbol, int len) | |
317 | { | |
318 | int i; | |
319 | ||
320 | for (i = 0; i < len - 1; i++) | |
b3dbb4ec | 321 | token_profit[ symbol[i] + (symbol[i + 1] << 8) ]++; |
1da177e4 LT |
322 | } |
323 | ||
324 | /* decrease the count for all the possible tokens in a symbol */ | |
325 | static void forget_symbol(unsigned char *symbol, int len) | |
326 | { | |
327 | int i; | |
328 | ||
329 | for (i = 0; i < len - 1; i++) | |
b3dbb4ec | 330 | token_profit[ symbol[i] + (symbol[i + 1] << 8) ]--; |
1da177e4 LT |
331 | } |
332 | ||
b3dbb4ec | 333 | /* remove all the invalid symbols from the table and do the initial token count */ |
1da177e4 LT |
334 | static void build_initial_tok_table(void) |
335 | { | |
b3dbb4ec | 336 | unsigned int i, pos; |
1da177e4 | 337 | |
b3dbb4ec PM |
338 | pos = 0; |
339 | for (i = 0; i < table_cnt; i++) { | |
1da177e4 | 340 | if ( symbol_valid(&table[i]) ) { |
b3dbb4ec PM |
341 | if (pos != i) |
342 | table[pos] = table[i]; | |
343 | learn_symbol(table[pos].sym, table[pos].len); | |
344 | pos++; | |
1da177e4 | 345 | } |
1da177e4 | 346 | } |
b3dbb4ec | 347 | table_cnt = pos; |
1da177e4 LT |
348 | } |
349 | ||
350 | /* replace a given token in all the valid symbols. Use the sampled symbols | |
351 | * to update the counts */ | |
b3dbb4ec | 352 | static void compress_symbols(unsigned char *str, int idx) |
1da177e4 | 353 | { |
b3dbb4ec PM |
354 | unsigned int i, len, size; |
355 | unsigned char *p1, *p2; | |
1da177e4 | 356 | |
b3dbb4ec | 357 | for (i = 0; i < table_cnt; i++) { |
1da177e4 LT |
358 | |
359 | len = table[i].len; | |
b3dbb4ec PM |
360 | p1 = table[i].sym; |
361 | ||
362 | /* find the token on the symbol */ | |
363 | p2 = memmem(p1, len, str, 2); | |
364 | if (!p2) continue; | |
365 | ||
366 | /* decrease the counts for this symbol's tokens */ | |
367 | forget_symbol(table[i].sym, len); | |
368 | ||
369 | size = len; | |
1da177e4 LT |
370 | |
371 | do { | |
b3dbb4ec PM |
372 | *p2 = idx; |
373 | p2++; | |
374 | size -= (p2 - p1); | |
375 | memmove(p2, p2 + 1, size); | |
376 | p1 = p2; | |
377 | len--; | |
378 | ||
379 | if (size < 2) break; | |
380 | ||
1da177e4 | 381 | /* find the token on the symbol */ |
b3dbb4ec | 382 | p2 = memmem(p1, size, str, 2); |
1da177e4 | 383 | |
b3dbb4ec | 384 | } while (p2); |
1da177e4 | 385 | |
b3dbb4ec | 386 | table[i].len = len; |
1da177e4 | 387 | |
b3dbb4ec PM |
388 | /* increase the counts for this symbol's new tokens */ |
389 | learn_symbol(table[i].sym, len); | |
1da177e4 LT |
390 | } |
391 | } | |
392 | ||
393 | /* search the token with the maximum profit */ | |
b3dbb4ec | 394 | static int find_best_token(void) |
1da177e4 | 395 | { |
b3dbb4ec | 396 | int i, best, bestprofit; |
1da177e4 LT |
397 | |
398 | bestprofit=-10000; | |
b3dbb4ec | 399 | best = 0; |
1da177e4 | 400 | |
b3dbb4ec PM |
401 | for (i = 0; i < 0x10000; i++) { |
402 | if (token_profit[i] > bestprofit) { | |
403 | best = i; | |
404 | bestprofit = token_profit[i]; | |
1da177e4 | 405 | } |
1da177e4 | 406 | } |
1da177e4 LT |
407 | return best; |
408 | } | |
409 | ||
410 | /* this is the core of the algorithm: calculate the "best" table */ | |
411 | static void optimize_result(void) | |
412 | { | |
b3dbb4ec | 413 | int i, best; |
1da177e4 LT |
414 | |
415 | /* using the '\0' symbol last allows compress_symbols to use standard | |
416 | * fast string functions */ | |
417 | for (i = 255; i >= 0; i--) { | |
418 | ||
419 | /* if this table slot is empty (it is not used by an actual | |
420 | * original char code */ | |
421 | if (!best_table_len[i]) { | |
422 | ||
423 | /* find the token with the breates profit value */ | |
424 | best = find_best_token(); | |
425 | ||
426 | /* place it in the "best" table */ | |
b3dbb4ec PM |
427 | best_table_len[i] = 2; |
428 | best_table[i][0] = best & 0xFF; | |
429 | best_table[i][1] = (best >> 8) & 0xFF; | |
1da177e4 LT |
430 | |
431 | /* replace this token in all the valid symbols */ | |
b3dbb4ec | 432 | compress_symbols(best_table[i], i); |
1da177e4 LT |
433 | } |
434 | } | |
435 | } | |
436 | ||
437 | /* start by placing the symbols that are actually used on the table */ | |
438 | static void insert_real_symbols_in_table(void) | |
439 | { | |
b3dbb4ec | 440 | unsigned int i, j, c; |
1da177e4 LT |
441 | |
442 | memset(best_table, 0, sizeof(best_table)); | |
443 | memset(best_table_len, 0, sizeof(best_table_len)); | |
444 | ||
b3dbb4ec PM |
445 | for (i = 0; i < table_cnt; i++) { |
446 | for (j = 0; j < table[i].len; j++) { | |
447 | c = table[i].sym[j]; | |
448 | best_table[c][0]=c; | |
449 | best_table_len[c]=1; | |
1da177e4 LT |
450 | } |
451 | } | |
452 | } | |
453 | ||
454 | static void optimize_token_table(void) | |
455 | { | |
1da177e4 LT |
456 | build_initial_tok_table(); |
457 | ||
458 | insert_real_symbols_in_table(); | |
459 | ||
41f11a4f | 460 | /* When valid symbol is not registered, exit to error */ |
b3dbb4ec | 461 | if (!table_cnt) { |
41f11a4f YS |
462 | fprintf(stderr, "No valid symbol.\n"); |
463 | exit(1); | |
464 | } | |
465 | ||
1da177e4 LT |
466 | optimize_result(); |
467 | } | |
468 | ||
469 | ||
b3dbb4ec | 470 | int main(int argc, char **argv) |
1da177e4 | 471 | { |
41f11a4f YS |
472 | if (argc >= 2) { |
473 | int i; | |
474 | for (i = 1; i < argc; i++) { | |
475 | if(strcmp(argv[i], "--all-symbols") == 0) | |
476 | all_symbols = 1; | |
477 | else if (strncmp(argv[i], "--symbol-prefix=", 16) == 0) { | |
478 | char *p = &argv[i][16]; | |
479 | /* skip quote */ | |
480 | if ((*p == '"' && *(p+2) == '"') || (*p == '\'' && *(p+2) == '\'')) | |
481 | p++; | |
482 | symbol_prefix_char = *p; | |
483 | } else | |
484 | usage(); | |
485 | } | |
486 | } else if (argc != 1) | |
1da177e4 LT |
487 | usage(); |
488 | ||
489 | read_map(stdin); | |
490 | optimize_token_table(); | |
491 | write_src(); | |
492 | ||
493 | return 0; | |
494 | } |